Smartphone App Recommendation System Using Usage Pattern Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face inconvenience in quickly accessing frequently used smartphone applications due to the large number of apps installed, as existing solutions do not effectively consider usage frequency, time, and situation-based app recommendations.
Innovation Solution
A method and apparatus that utilize artificial intelligence to learn app usage patterns, collecting data on frequency and time of use, and recommend frequently used apps based on specific conditions such as daily time periods and situations, preventing overlap between different situation-based app lists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If apps are organized in folders based on frequency of use, alphabetical order, or type of service, then apps can be categorized and stored, but the time and effort required for users to search for and execute desired apps increases
Solution Approach 1:
The system performs preliminary analysis of app usage patterns, frequency, and temporal characteristics before the user needs to access an app. By pre-processing usage data and identifying frequently used apps in advance, the system prepares personalized shortcuts or recommendations that are ready for immediate presentation, eliminating the need for users to search through folders at the moment of need.
Solution Approach 2:
The patent replaces manual folder-based organization with an automated intelligent system that uses machine learning algorithms to analyze usage patterns and automatically generate personalized app recommendations. This substitutes the mechanical sorting approach with an adaptive computational system that continuously learns and improves based on user behavior.
2Adaptability or versatility
If probability-based situation estimation is used to recommend apps, then app recommendations can be provided based on estimated situations, but factors such as number of times of execution, usage time, and usage pattern are not considered
Solution Approach 1:
The system merges multiple dimensions of app usage information including frequency of execution, total usage time, temporal patterns, and situational context into a unified analysis framework. By combining these diverse data sources, the system creates a comprehensive view of app usage that enables accurate identification of frequently used apps while maintaining adaptability to different situations.
Solution Approach 2:
The system implements feedback mechanisms that continuously monitor actual app usage and compare it with recommendations. This feedback loop allows the system to refine its understanding of usage patterns over time, improving the precision of frequently used app extraction while maintaining situational adaptability through continuous learning.
Data Source
AI summary
The present disclosure provides a method of displaying frequently used applications and an apparatus using the same. The method of displaying frequently used applications includes: collecting application usage information in at least any one of a time period of a day or a user activity; determining an application usage pattern based on the collected application usage information; and determining frequently used applications in the time period of a day or the user activity based on the determined application usage pattern, and displaying the determined frequently used applications.